Fully Automated Organ Segmentation in Male Pelvic CT Images
This work addresses the need for accurate organ segmentation in prostate cancer radiotherapy treatment planning, representing an incremental improvement with specific gains in automation and performance.
The authors tackled the problem of segmenting prostate and surrounding organs in male pelvic CT images for radiotherapy planning, achieving mean Dice coefficients of 90% for prostate, 96% for left femoral head, 95% for right femoral head, 95% for bladder, and 84% for rectum using a fully automated deep learning method.
Accurate segmentation of prostate and surrounding organs at risk is important for prostate cancer radiotherapy treatment planning. We present a fully automated workflow for male pelvic CT image segmentation using deep learning. The architecture consists of a 2D localization network followed by a 3D segmentation network for volumetric segmentation of prostate, bladder, rectum, and femoral heads. We used a multi-channel 2D U-Net followed by a 3D U-Net with encoding arm modified with aggregated residual networks, known as ResNeXt. The models were trained and tested on a pelvic CT image dataset comprising 136 patients. Test results show that 3D U-Net based segmentation achieves mean (SD) Dice coefficient values of 90 (2.0)% ,96 (3.0)%, 95 (1.3)%, 95 (1.5)%, and 84 (3.7)% for prostate, left femoral head, right femoral head, bladder, and rectum, respectively, using the proposed fully automated segmentation method.